U.S. patent number 7,606,419 [Application Number 11/213,784] was granted by the patent office on 2009-10-20 for translated document image production device, recording medium and translated document image production method.
This patent grant is currently assigned to Fuji Xerox Co., Ltd.. Invention is credited to Shunichi Kimura, Yutaka Koshi.
United States Patent |
7,606,419 |
Koshi , et al. |
October 20, 2009 |
Translated document image production device, recording medium and
translated document image production method
Abstract
A translated document image production device has an extraction
section, a translation section and a region backfill section. The
extraction section extracts a group of pixels to be processed, the
group of pixels having pixels for constituting a character or a
character string. The translation section translates the character
or character string. The region backfill section backfills the
extracted group of pixels to maintain the background of the
translated character or character string and that writes the
translated character and character string.
Inventors: |
Koshi; Yutaka
(Ashigarakami-gun, JP), Kimura; Shunichi
(Ashigarakami-gun, JP) |
Assignee: |
Fuji Xerox Co., Ltd. (Tokyo,
JP)
|
Family
ID: |
36970975 |
Appl.
No.: |
11/213,784 |
Filed: |
August 30, 2005 |
Prior Publication Data
|
|
|
|
Document
Identifier |
Publication Date |
|
US 20060204111 A1 |
Sep 14, 2006 |
|
Foreign Application Priority Data
|
|
|
|
|
Mar 8, 2005 [JP] |
|
|
2005-064130 |
|
Current U.S.
Class: |
382/176; 704/2;
382/276; 382/229; 382/182 |
Current CPC
Class: |
G06T
11/60 (20130101); G06F 40/58 (20200101); G06T
7/11 (20170101); G06K 2209/015 (20130101) |
Current International
Class: |
G06K
9/34 (20060101); G06K 9/18 (20060101); G06K
9/36 (20060101); G06K 9/72 (20060101) |
Field of
Search: |
;382/276,176-179,182,185,186,229 ;704/2-7 |
References Cited
[Referenced By]
U.S. Patent Documents
Foreign Patent Documents
|
|
|
|
|
|
|
A 5-274397 |
|
Oct 1993 |
|
JP |
|
A 2002-108855 |
|
Apr 2002 |
|
JP |
|
Primary Examiner: Carter; Aaron W
Attorney, Agent or Firm: Oliff & Berridge, PLC
Claims
What is claimed is:
1. A translated document image production device, comprising: an
extraction section of the device that extracts a group of pixels
constituting only a foreground of a character or a character string
to be processed from pixels constituting a background of the
character or character string; a translation section of the device
that translates the character or character string into a translated
character or character string; and a region backfill section of the
device that backfills only the extracted group of pixels with pixel
values representing the background of the character or character
string and that writes the translated character or character string
over the background.
2. The translated document image production device according to
claim 1, wherein, the region backfill section backfills the
extracted group of pixels by replacing the extracted group of
pixels with pixels having a same pixel value as pixels in a
vicinity of the extracted pixels.
3. A translated document image production method, comprising:
extracting a group of pixels constituting only a foreground of a
character or a character string to be processed from pixels
constituting a background of the character or character string;
translating the character or character string into a translated
character or character string; and backfilling only the extracted
group of pixels with pixel values representing the background of
the character or character string and that writes the translated
character or character string over the background, wherein the
method is implemented by a computer.
4. The translated document image production method according to
claim 3, wherein, the extracted group of pixels is backfilled by
replacing the extracted group of pixels with pixels having a same
pixel value as pixels in a vicinity of the extracted pixels.
5. A storage medium readable by a computer, the storage medium
storing a program of instructions executable by the computer to
perform a function for producing a translated document, the
function comprising: extracting a group of pixels constituting only
a foreground of a character or a character string to be processed
from pixels constituting a background of the character or character
string; translating the character or character string into a
translated character or character string; and backfilling only the
extracted group of pixels with pixel values representing the
background of the character or character string and that writes the
translated character or character string over the background.
6. The storage medium according to claim 5, wherein, the extracted
group of pixels is backfilled by replacing the extracted group of
pixels with pixels having a same pixel value as pixels in a
vicinity of the extracted pixels.
7. A translated document image production device, comprising: an
input section that inputs a bitmap image obtained by scanning a
document; an extraction section that specifies a region from the
input bitmap image where a character or character string is
inscribed, and a group of pixels constituting only a foreground of
a character or a character string to be processed from pixels
constituting a background of the character or character string; a
character discrimination section that discriminates the character
or character string displayed by the extracted group of pixels; a
translation section that translates the discriminated character or
character string, and obtains a new group of pixels that draw a new
character or character string obtained by the translation; a region
backfill section that replaces only the extracted group of pixels
with other pixels obtained by performing predetermined
interpolation processing, matching the value of pixels constituting
the character or character string in the specified region, with a
value of pixels constituting the background in the vicinity of
those pixels; an image writing section that obtains a bitmap image
of a translated document by writing the new group of pixels to the
region in which the pixels have been replaced; and an output
section that outputs the obtained bitmap image of the translated
document, wherein the sections are implemented using hardware.
8. The translated document image production device according to
claim 7, wherein the region backfill section replaces the extracted
group of pixels with pixels having a same pixel value as pixels in
a vicinity of the extracted pixels.
9. A computer readable recording medium recording a program for
causing a computer to execute: inputting a bitmap image obtained by
scanning a document; specifying a region from the input bitmap
image where a character or character string is inscribed, and
extracting a group of pixels constituting only a foreground of a
character or a character string to be processed from pixels
constituting a background of the character or character string;
discriminating the character or character string displayed by the
extracted group of pixels; translating the discriminated character
or character string, and obtaining a new group of pixels that draw
a new character or character string obtained by the translation;
replacing only the extracted group of pixels with other pixels
obtained by performing predetermined interpolation processing,
matching a value of the pixels constituting the character or
character string in the specified region, with a value of pixels
constituting the background in a vicinity of those pixels;
obtaining a bitmap image of a translated document by writing the
new group of pixels to the region in which the pixels have been
replaced; and outputting the obtained bitmap image of the
translated document.
10. The recording medium according to claim 9, wherein the
replacing step comprises replacing the extracted group of pixels
with pixels having a same pixel value as pixels in a vicinity of
the extracted pixels.
Description
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to technologies for translating a
bitmapped document.
2. Description of the Related Art
Various technologies have been proposed up to the present in order
to scan a document provided on a paper medium and optimally handle
the document as document data. For example, Japanese Patent
Application Laid-Open Publication H05-274397 discloses technology
that performs character discrimination in a scanned document, and
after appropriately modifying the contents of that data, exchanges
it between remote devices. Also, Japanese Patent Application
Laid-Open Publication 2002-108855 discloses technology that
recognizes characters with high accuracy, even from a portion of a
document that includes maps, illustrations, or the like.
In the past, attempts have been made to replace only the character
portion of a document with new characters translated into another
language, while maintaining the appearance of colors, patterns, and
the like that form the background of the document, by altering the
document data, obtained by scanning the document, at the pixel
level. Ordinarily, document data is provided as a bitmap wherein
the contents of the scanned document are displayed by pairs of
pixel location and pixel value. Accordingly, when a new character
obtained by translating a character inscribed in that region is
overwritten as-is in the region within the bitmap, as shown in FIG.
4 (A), because the overwriting character lies on top of the
originally inscribed character, its legibility is significantly
compromised. On the other hand, it is certainly possible to avoid a
piling of characters if the new characters are written after the
pixel value of the pixels constituting the originally present
character is set once to white. However, even when adopting this
sort of method, when the background of the document is drawn in a
color or pattern other than white, as shown in FIG. 4 (B), a white
outline of the character remains in the background of the document,
compromising legibility in the same manner as before.
The present invention has been made in view of the above
circumstances, and provides an arrangement with which it is
possible to replace only the character portion of a document with
new characters translated into another language, while maintaining
the appearance of colors, patterns, and the like that form the
background of the document.
SUMMARY OF THE INVENTION
According to an aspect of the present invention, there is provided
a translated document image production device. The translated
document image production device has an extraction section, a
translation section and a region backfill section. The extraction
section extracts a group of pixels to be processed, the group of
pixels having pixels for constituting a character or a character
string. The translation section translates the character or
character string. The region backfill section backfills the
extracted group of pixels to maintain the background of the
translated character or character string and that writes the
translated character and character string.
According to an embodiment of the present invention, it is possible
to replace only the character portion of a document with new
characters translated into another language, while maintaining the
appearance of colors, patterns, and the like that form the
background of the document.
BRIEF DESCRIPTION OF THE DRAWINGS
Embodiments of the present invention will be described in detail
based on the following figures, wherein:
FIG. 1 is a hardware configuration diagram of the translated
document image production device;
FIGS. 2(A), 2(B), 2(C), 2(D) and 2(E) are drawings that show the
transformation of a bitmap image;
FIG. 3 is a hardware configuration diagram of the translated
document image production device (in a modified embodiment);
and
FIGS. 4(A) and 4(B) are drawings that illustrate the conventional
technology.
DETAILED DESCRIPTION OF THE INVENTION
Embodiment of the Invention
Below follows a description of an embodiment according to the
present invention.
In the present embodiment, after discriminating a character from a
bitmapped image obtained by scanning a document wherein text is
printed in black on paper of a single color, a new translated
character is written after backfilling the pixels constituting that
character with pixels that have the same attributes as the
paper.
FIG. 1 is a block diagram that shows the hardware configuration of
the translated document image production device 1 according to the
present embodiment. As shown in FIG. 1, the translated document
image production device 1 is provided with a document image data
input portion 10, a character portion image extraction portion 11,
a character discrimination portion 12, a translation portion 13, a
character image data production portion 14, a background image
extraction portion 15, a character portion image deletion portion
16, a backfill portion 17, an image writing portion 18, and a
document image data output portion 19.
The document image data input portion 10 inputs a bitmap image of
the document into the device 1 itself The bitmap image inputted via
the document image data input portion 10 has pairs including an
address that indicates the position of the pixels constituting the
document image and an RGB value that expresses the attributes of
those pixels as an RGB 256 gradation. For example, an RGB value of
(0, 0, 0) is shown when the pixel at a particular address is white,
and an RGB value of (255, 255, 255) is shown when the pixel at a
particular address is black.
The character portion image extraction portion 11 specifies one
rectangular region from among various inscribed rectangular areas
(hereinafter, a rectangular region to be processed will be referred
to as a "processing target region") in the bitmap image to be
processed that is input via the document image data input portion
10, and supplies position information that indicates the position
of the specified processing target region to the background image
extraction portion 15 and the character portion image deletion
portion 16. Also, this character portion image extraction portion
11 extracts the group of pixels constituting the character in the
processing target region. Based on the group of pixels extracted by
the character portion image extraction portion 11, the character
discrimination portion 12 identifies the character displayed by
that group of pixels, and supplies the character code of the
discriminated character to the translation portion 13. The
translation portion 13 translates the character code supplied from
the character discrimination portion 12 and obtains a new character
code, and supplies the newly obtained character code to the
character image data production portion 14. For example, when the
character code of the Latin character "A" is supplied from the
character discrimination portion 12, the character code of the
Japanese hiragana character is supplied to the character image data
production portion 14. The character image data production portion
14 produces image data corresponding to the character code supplied
from the character discrimination portion 12, and supplies that
character image to the image writing portion 18.
After specifying from the bitmap image the processing target region
indicated by the position information supplied from the character
portion image extraction portion 11, the background image
extraction portion 15 extracts the RGB values of the pixels
constituting the background of the character within that
region.
On the other hand, after specifying from the bitmap image the
processing target region indicated by the position information
supplied from the character portion image extraction portion 11,
the character portion image deletion portion 16 resets the RGB
values of the pixels constituting the background of the character
within that region to white (0, 0, 0). Then, the backfill portion
17 newly associates the same RGB values as those extracted by the
background image extraction portion 15 with the addresses of the
pixels reset by the character portion image deletion portion 16.
Therefore, the pixels that constituted the character in the
processing target region are replaced by pixels having the same RGB
value as the pixels that constituted the background, and the area
within the processing target region is concealed by pixels having
the same color as the background.
In the processing target region that has been processed by the
backfill portion 17, the image writing portion 18 fills the
character image supplied from the character image data production
portion 14.
When a bitmap image containing multiple characters is input from
the document image data input portion 10, the region where each of
those characters is inscribed is made a processing target region,
and the processing performed by the processing portions described
above is repeated. Thus, a new bitmap image is obtained that
displays the translated document. The newly obtained bitmap image
is output to an external printing device via the document image
data output portion 19.
Transformation of the document bitmap image realized by the
processing performed by the processing portions described above is
specifically explained here with reference to FIG. 2 as an example.
FIG. 2 (A) shows a bitmap image that has been input from the
document image data input portion 10. As shown in FIG. 2 (A), in
this bitmap image the Latin character "A" is drawn by the group of
black pixels (RGB value: 255, 255, 255), and the background of this
character is drawn by blue pixels (RGB value: 0, 255, 0). FIG. 2
(B) shows the bitmap image after being processed by the character
portion image extraction portion 11. At this stage, the rectangular
region in which the Latin character "A" is inscribed is specified
as a processing target region X, and the character code of the
character "A" shown by the group of black pixels within that
processing target region X is supplied from the character
discrimination portion 12 to the translation portion 13.
FIG. 2 (C) shows the bitmap image after being processed by the
character portion image deletion portion 16. At this stage, by
setting the RGB value of the pixels constituting the Latin
character "A" to white, the character "A" is cut out in white from
the processing target region X.
FIG. 2 (D) shows the bitmap image after being processed by the
backfill portion 17. At this stage, by newly associating the
background RGB value extracted by the background image extraction
portion 15 with the addresses of the reset pixels, the entire
processing target region X is concealed by the background
color.
FIG. 2 (E) shows the bitmap image after being processed by the
image writing portion 18. In this stage, the image of the Japanese
hiragana character obtained by translation is newly written to
approximately the center position of the processing target region
X. This newly written character is produced by the character image
data production portion 14 based on the character code supplied by
the translation portion 13.
The translated document image production device 1 described above,
after discriminating a character from the processing target region
of the input bitmap image, writes a new character image obtained by
translation after backfilling the pixels of the region in which
that character is inscribed with the same color as the background.
Accordingly, it is possible to avoid compromising the legibility of
the written character due to some influence from the background
color of the source document.
Modified Embodiments
Various modified embodiments of the invention are also
possible.
The embodiment described above is constructed supposing a
circumstance in which the background of the character in the input
bitmap image is a single color, and is not suitable for translation
of documents in which a pattern is pre-printed on paper. On the
other hand, by furnishing the backfill portion 17 with a function
as described below, it is possible to apply the present invention
even to documents having a background that is not a single color.
The backfill method described as a modified embodiment includes
copying the processing target region and a separated background
portion as a patch image, and after adjusting the size of the patch
to an appropriate size, pasting it to the processing target region.
In the present invention, any method may be adopted for that
backfill method as is necessary, if the backfill portion 17 is
furnished with a function that performs predetermined interpolation
processing, matching the group of pixels extracted from the
processing target region with the drawn content of pixel groups in
the vicinity of the extracted group of pixels.
In the embodiment described above, after the character portion
image deletion portion 16 sets the RGB value of the pixels
constituting the character in the processing target region to
white, the backfill portion 17 newly associates the same RGB values
as those extracted by the background image extraction portion 15
with the addresses of the reset pixels. However, this resetting by
the character portion image deletion portion 16 is not a required
process in the present invention; the RGB value of the pixels
constituting the character may be immediately replaced with the
same RGB value as that extracted by the background image extraction
portion 15.
In the embodiment described above, the backfill portion 17 newly
associated the RGB value of the background extracted by the
background image extraction portion 15 with the addresses of the
pixels reset by the character portion image deletion portion 16.
That is, the RGB value of the pixels backfilled by the backfill
portion 17 was determined depending on the results of the
extraction performed by the background image extraction portion 15.
On the other hand, by pre-setting the RGB value backfilled by the
backfill portion 17 to a default value indicating the primary
colors white (RGB value: 0, 0, 0) and blue (RGB value: 0, 255, 0),
a simple configuration may be adopted wherein it is not necessary
to furnish the background image extraction portion 15.
FIG. 3 is a block diagram that shows the hardware configuration of
the translated document image production device 1 according to this
modified embodiment. As shown in FIG. 3, the translated document
image production device 1 according to this modified embodiment is
provided with a document image data input portion 10, a character
portion image extraction portion 11, a character discrimination
portion 12, a translation portion 13, a character image data
production portion 14, a character portion image deletion portion
16, a backfill portion 17, an image writing portion 18, and a
document image data output portion 19, but unlike the translated
document image production device 1 shown in FIG. 1, it does not
include a background image extraction portion 15. Instead, in the
backfill portion 17 in the present modified embodiment, a
predetermined RGB value is set in advance for the pixels that
should be backfilled as the background, and the backfill portion 17
newly associates this set RGB value with the address of the pixels
reset by the character portion image deletion portion 16. Even with
this sort of simple configuration, it is possible to realize the
operation of the present invention.
The embodiment described above is configured with the translation
portion 13 and the character image data production portion 14 as
separate portions, but it is also possible to incorporate the
function of the character image data production portion 14 into the
translation portion 13. In the case of this modified embodiment,
the translation portion 13 itself produces a new pixel group that
displays the character indicated by the character code obtained by
translation, and supplies it to the image writing portion 18.
In the embodiment described above, only one character is inscribed
in a single processing target region, but it is also possible to
make the processing target region a rectangular region including a
character string for each predetermined group.
As described above, the present invention provides, in one aspect,
a translated document image production device, having: an
extraction section that extracts a group of pixels to be processed,
the group of pixels having pixels for constituting a character or a
character string; a translation section that translates the
character or character string; and a region backfill section that
backfills the extracted group of pixels to maintain the background
of the translated character or character string and that writes the
translated character and character string.
In this aspect, the region backfill section may backfill the
extracted group of pixels by replacing the extracted group of
pixels with pixels having the same pixel value as pixels in the
vicinity of the extracted pixels.
Another aspect of the present invention is a translated document
image production method, having: extracting a group of pixels to be
processed, the group of pixels having pixels for constituting a
character or a character string; translating the character or
character string; and backfilling the extracted group of pixels to
maintain the background of the translated character or character
string and that writes the translated character and character
string.
In this aspect, the extracted group of pixels may be backfilled by
replacing the extracted group of pixels with pixels having the same
pixel value as pixels in the vicinity of the extracted pixels.
Another aspect of the present invention is a storage medium
readable by a computer, the storage medium storing a program of
instructions executable by the computer to perform a function for
producing a translated document, the function having: extracting a
group of pixels to be processed, the group of pixels having pixels
for constituting a character or a character string; translating the
character or character string; and backfilling the extracted group
of pixels to maintain the background of the translated character or
character string and that writes the translated character and
character string.
In this aspect, the extracted group of pixels may be backfilled by
replacing the extracted group of pixels with pixels having the same
pixel value as pixels in the vicinity of the extracted pixels.
Another aspect of the present invention is a translated document
image production device, having: an input section that inputs a
bitmap image obtained by scanning a document; an extraction section
that specifies a region from the input bitmap image where a
character or character string is inscribed, and extracts a group of
pixels constituting the character or character string in the
specified region; a character discrimination section that
discriminates the character or character string displayed by the
extracted group of pixels; a translation section that translates
the discriminated character or character string, and obtains a new
group of pixels that draw a new character or character string
obtained by the translation; a region backfill section that
replaces the extracted group of pixels with other pixels obtained
by performing predetermined interpolation processing, matching the
pixels constituting the character or character string in the
specified region, with pixels constituting the background in the
vicinity of those pixels; an image writing section that obtains a
bitmap image of a translated document by writing the newly obtained
group of pixels to the region in which the pixels have been
replaced; and an output section that outputs the obtained bitmap
image of the translated document.
In this aspect, the region backfill section may replace the
extracted group of pixels with pixels having the same pixel value
as pixels in the vicinity of the extracted pixels.
Another aspect of the present invention is a computer readable
recording medium recording a program for causing a computer to
execute: inputting a bitmap image obtained by scanning a document;
specifying a region from the input bitmap image where a character
or character string is inscribed, and extracting a group of pixels
constituting the character or character string in the specified
region; discriminating the character or character string displayed
by the extracted group of pixels; translating the discriminated
character or character string, and obtaining a new group of pixels
that draw a new character or character string obtained by the
translation; replacing the extracted group of pixels with other
pixels obtained by performing predetermined interpolation
processing, matching the pixels constituting the character or
character string in the specified region, with pixels constituting
the background in the vicinity of those pixels; obtaining a bitmap
image of a translated document by writing the newly obtained group
of pixels to the region in which the pixels have been replaced; and
outputting the obtained bitmap image of the translated
document.
In this aspect, the replacing step may have replacing the extracted
group of pixels with pixels having the same pixel value as pixels
in the vicinity of the extracted pixels.
The foregoing description of the embodiments of the present
invention has been provided for the purposes of illustration and
description. It is not intended to be exhaustive or to limit the
invention to the precise forms disclosed. Obviously, many
modifications and variations will be apparent to practitioners
skilled in the art. The embodiments were chosen and described in
order to best explain the principles of the invention and its
practical applications, thereby enabling others skilled in the art
to understand the invention for various embodiments and with the
various modifications as are suited to the particular use
contemplated. It is intended that the scope of the invention be
defined by the following claims and their equivalents.
The entire disclosure of Japanese Patent Application No.
2005-064130 filed on Mar. 8, 2005 including specification, claims,
drawings and abstract is incorporated herein by reference in its
entirety.
* * * * *